Low-Energy Ionization “Cooling” David Neuffer Fermilab

2 Outline  Low-energy cooling-protons  ions and “beta-beams”  Low-energy cooling – muons  emittance exchange

3 μ Cooling Regimes  Efficient cooling requires:  Frictional Cooling (<1MeV/c) Σ g =~3  Ionization Cooling (~0.3GeV/c) Σ g =~2  Radiative Cooling (>1TeV/c) Σ g =~4  Low-ε t cooling Σ g =~2β 2 (longitudinal heating)

4 Cooling/Heating equations  Cooling equations are same as used for muons  mass = a m p, charge = z e  Some formulae may be inaccurate for small β=v/c  Add heating through nuclear interactions  Ionization/recombination should be included  For small β, longitudinal dE/dx heating is large  At β =0.1, g L = -1.64, ∑ g = 0.36  Coupling only with x cannot obtain damping in both x and z 

5 Low-energy “cooling” of ions  Ionization cooling of protons/ ions has been unattractive because nuclear reaction rate is competitive with energy-loss cooling rate  And other cooling methods are available  But can have some value if the goal is beam storage to obtain nuclear reactions  Absorber is also nuclear interaction medium  Y. Mori – neutron beam source –NIM paper  C. Rubbia, Ferrari, Kadi, Vlachoudis – source of ions for β-beams

6 Miscellaneous Cooling equations For small β: Better for larger mass?

7 Example  ERIT-P-storage ring to obtain directed neutron beam (Mori-Okabe, FFAG05)  10 MeV protons  9 Be target for neutrons  σ ≈ 0.5 barns  β = v/c =0.145  Large δE heating  Baseline Absorber  5µ Be absorber  δE p =~36 keV/turn  Design Intensity  1000Hz, 6.5×10 10 p/cycle  100W primary beam  < 1.5 kW on foil –0.4 kW at n turns =1000

8 ERIT results  With only production reaction, lifetime is turns  With baseline parameters, cannot cool both x and E  Optimal x-E exchange increases storage time from 1000 to 3000 turns (3850 turns = 1ms)  With x-y-E coupling, could cool 3-D with g i = 0.12  Cooling time would be ~5000 turns  With β  =0.2m,  δE rms = 0.4MeV,  ,N = m (x rms =2.3cm)  x rms = 7.3cm at β  =2m (would need r =20cm arc apertures) (but 1ms refill time would make this unnecessary)

9 ERIT-recent results  Lattice changed from spiral to radial sector  spiral sector had too small vertical aperture  With cooling effects, beam has ~1000 turn lifetime in ICOOL simulation (Mori and Okabe FFAG06)

10 Emittance exchange parameters  with g L,0 = -1.63, η = 0.5m,  need G = 3.5 m -1 to get g L =0.12 –(L W =0.3m)  For 3-D cooling, need to mix with both x and y  Solenoid cooling rings  Also “Moebius” lattice (R. Talman)  Single turn includes x-y exchange transport –solenoid(s) or skew quads –in zero dispersion region for simplicity –Solenoid: BL = π Bρ –For 10 MeV p : BL = 1.44T-m  Complete period is 2 turns   Beam Center Geometry of wedge absorber L W =1/G

11 “Wedge” for thin foil  Obtain variable thickness by bent foil (Mori et al.)  Choose x 0 =0.15m, L W =0.3, δ o =5µ  then a=0.15, δ R =2.5  for g L,0 = -1.63, η = 0.5m  Barely Compatible with β * = 0.2m  Beam energy loss not too large?  <1kW Power on foil a=0.15 Beam

12 Other heating terms:  Mixing of transverse heating with longitudinal could be larger effect: (Wang & Kim) At ERIT parameters: β x = 1.0m, β z = 16m, η=0.6m, Be absorber, dδ 2 /ds= , d θ 2 /ds= only 5%, 1.5% changes … At β x = 0.2m, ~25% change …

13 β-beam Scenario Neutrino Source Decay Ring Ion production: Target - Ion ( 8 B or 8 Li ?) Conversion Ring Ion Driver 25MeV Li ? Decay ring B  = 400 Tm B = 5 T C = 3300 m L ss = 1100 m 8 B:  = 80 8 Li:  = 50 Acceleration to medium energy RCS 8z GeV/c Acceleration to final energy Fermilab Main Injector Experiment Ion acceleration Linac Beam preparation ECR pulsed Ion productionAccelerationNeutrino source

14 Conventional Beta beam ion source ν-sources v * -sources  Want:  lifetime ~1s  large ν-energy  ν and ν*  easily extracted atoms  Number of possible ions is limited (v sources easier)  Noble gases easier to extract  6 He 2 “easiest” E ν* =1.94MeV  18 Ne 10 for ν E ν =1.52MeV  ( 8 B, 8 Li) have E ν,ν * =~7MeV  Want ν and ν* / year…

15 β-beam Scenario (Rubbia et al.)  Produce Li and inject at 25 MeV  Charge exchange injection  nuclear interaction at gas jet target produces 8 Li or 8 B  Multiturn with cooling maximizes ion production  8 Li or 8 B is caught on stopper(W)  heated to reemit as gas  8 Li or 8 B gas is ion source for β- beam accelerate  Accelerate to Bρ = 400 T-m  Fermilab main injector  Stack in storage ring for:  8 B  8 Be + e + +ν or 8 Li  8 Be + e - + ν* neutrino source

16 Cooling for β -beams (Rubbia et al.-NuFACT06)  β-beam requires ions with appropriate nuclear decay  8 B  8 Be + e + + ν  8 Li  8 Be + e - + ν*  Ions are produced by nuclear interactions  6 Li + 3 He  8 B + n  7 Li + 2 H  8 Li + 1 H  Secondary ions must be collected and reaccelerated  Either heavy or light ion could be beam or target  Ref. 1 prefers heavy ion beam – ions are produced more forward (“reverse kinematics”)  He or 2 H beam on Li has other advantages ….  Parameters can be chosen such that target “cools” beam  (losses and heating from nuclear interactions, however…)

17 β-beams example: 6 Li + 3 He  8 B + n  Beam: 25MeV 6 Li +++  P Li =529.9 MeV/c Bρ = 0.59 T-m; v/c=  Absorber: 3 He  Z=2, A=3, I=31eV, z=3, a=6  dE/ds= 1180 MeV/gm/cm 2, L R = 70.9 gm/cm 2 (ρ He-3 = gm/cm 3 ) Liquid, (ρ He-3 = 0.134∙10 -3 P gm/cm3/atm in gas)  If g x = (Σ g = 0.37), β ┴ =0.3m at absorber  ε N,eq = ~ m-rad  σ x,rms = 1.2 cm at β ┴ =0.3m,  σ x,rms = 3.14 cm at β ┴ =2.0m  σ E,eq is ~ 0.4 MeV  ln[ ]=5.68

18 Cooling time/power : 6 Li + 3 He  8 B + n  Nuclear cross section for beam loss is 1 barn ( cm 2 ) or more  σ = cm 2, corresponds to ~5gm/cm 2 of 3 He  ~10 3-D cooling e-foldings …  Cross-section for 8 B production is ~10 mbarn  At best, of 6 Li is converted  Goal is /s of 8 B production  then at least Li 6 /s needed  Space charge limit is ~ Li/ring  Cycle time is <10 -3 s  If C=10m, τ =355 ns, 2820 turns/ms  5/2820=1.773*10 -3 gm/cm 2 (0.019 liquid density …)  2.1 MeV/turn energy loss and regain required …(0.7MV rf)  MW cooling rf power…

19 Complementary case- 7 Li + 2 H  8 Li + 1 H  Nuclear cross section for beam loss is 1 barn ( cm 2 ) or more  σ = cm 2, corresponds to ~3.3gm/cm 2 of 2 H  ~9 3-D cooling e-foldings …  Cross-section for 8 Li production is ~100 mbarn  of 7 Li is converted ?? 10 × better than 8 B neutrinos  Goal is /s of 8 Li production  then at least Li 7 /s needed  Space charge limit is ~ Li/ring  Cycle time can be up to s, but use s  If C=10m, τ =355 ns, 2820 turns  3.3/2820=1.2*10 -3 gm/cm 2 (0.007 liquid D density …)  1.3 MeV/turn energy loss and regain required …(0.43MV rf)  0.06 MW cooling rf power…

20 Space charge – Direct/inverse ?  At N = 10 12, B F =0.2, β = 0.094, z=3, a=6, ε N,rms = : δν ≈ 0.2  tolerable ??  Space charge sets limit on number of particles in beam and on transverse emittance  Effect is reduced for “direct kinematics”  (D/He beam, Li target)  Is “direct” source better than “inverse” source?  6 Li 3 beam + 3 He 2 target  or  3 He 2 beam + 6 Li 3 target  Beam energy, power on target less (1/2to 1/3)  Li foil or gas-jet target? Gas-jet nozzle for wedge effect  > 0.1MW power on target

21 Rubbia et al. not completely wrong  …But contains mistakes  Longitudinal emittance growth 2× larger  (than NuFACT06 presentation)  synchrotron oscillations reduce energy spread growth rate but not emittance growth rate  Emittance exchange needs x-y coupling and balancing of cooling rates to get 3-D cooling  More complicated lattice  3-D cooling needed to get enough ions  Increases equilibrium emittance, beam size  increase needed for space charge, however  Ion production to storage ring efficiency is not 100% …

22 µ Low-Energy “cooling”- emittance exchange  dP μ /ds varies as ~1/β 3  “Cooling” distance becomes very short: for liquid H at P μ = 10MeV/c  Focusing can get quite strong:  Solenoid:  β  =0.002m at 30T, 10MeV/c  ε N,eq = 1.5×10 -4 cm at 10MeV/c  Small enough for “low-emittance” collider PµPµ L cool 200 MeV/c cm 100 cm

23 ICOOL Simulation results  Low-Energy muons in H 2 absorber  50 MeV/c (4 cm H 2 )  30 MeV/c (0.7cm H 2 )  15 MeV/c (0.8 mm H 2 or 80 μ Be or… )  Could use gas absorbers/jets ?  Results follow rms eqns  less multiple scattering …  Typical section:  reduces P by 1/3P  δp increases by factor of 2  ε x, ε y reduced by 1/√2

24 ICOOL Multiple Scattering effects  New Model 6 (Fano model) much less rms scattering than Model 4(Moliere/Bethe)  At 200 MeV/c μ on H 2, M4 scattering ~10% > rms eq.  M6 scattering (θ 2 ) is ~30% less than rms eq.  Low energy scattering less at low momentum  At 15 MeV/c, M4 scattering ~40% < than rms eq.  M6 scattering (θ 2 ) is ~60% less than rms eq.  Which is more accurate? rms eq., model 4 or 6 or ??

25 Comments  Can fit into end-stage cooling (with similar effects?)  Can use gas jet absorbers to avoid having windows  P jet > 1 atm possible  Need “rf” to reduce dp/p (longer bunches for multistep )  1mm bunch can grow to 1m bunch length  Voltage is relatively small  L μ = 660βγ  Reacceleration of ~1m bunches Approximate effect Of low-E emittance exchange

26 Summary  Low energy ionization cooling has possible important applications  Protons for neutron generation (Mori et al.)  β-beam source production (Rubbia et al.)  Cooling of μ’s to minimum transverse emittance –REMEX that might work …  “Cooling” is predominantly emittance exchange  X-y-z exchange needed for “real” cooling

27 ILC Status

28 X-sections, kinematics